llama.cpp Introduces Decision Models and `/v1/systemone` Endpoint

Core summary

One sentence to understand this update

Georgi Gerganov announced that decision models are now available in llama.cpp via the `/v1/systemone` endpoint, enabling local, efficient, and private Jev-style inference.

Impact & opportunity

What this could mean

This feature empowers builders to implement advanced decision-making logic directly within local llama.cpp environments, enhancing privacy and performance for AI applications.